Difficulty in producing fluent speech is the core of stuttering, a neurodevelopmental speech disorder affecting children into adulthood with long-term psychosocial and economic impact. A characteristic of the disorder is that even when producing perceptually fluent speech, there are differences in the brain and behavioural processes for individuals who stutter compared to those that do not. Unfortunately, after decades of research involving behavioral description, structural and functional neuroimaging and computational modeling, no coherent neurobiological framework for the disorder exists. The current proposal is a first step toward developing a neural mechanistic model for stuttering with application to other neurodevelopmental disorders of spoken language production. The over-arching hypothesis is that stuttering is associated with a reduced level of neural inhibition, as reflected in a reduction in the negative blood-oxygen-level-dependent (BOLD) signal response (NBR). This reduction in inhibition during development has long-term and wide-ranging effects on brain structure and function and the stability and dynamics of neural networks for speech potentially impacting all or some sensorimotor, linguistic and cognitive behaviours. The empirical data and conceptual framework will impact speech production models and theories by adding a currently missing neural component process to the manner in which speech is produced and pave the way for a more informed and neurobiological model of neurodevelopmental speech disorders.